Gary M. Mawe PhD, Given D403, Department of Anatomy and Neurobiology, University of Vermont, Burlington, VT 05405, USA. Tel: (802) 656-8257; fax: (802) 656-8704; e-mail: email@example.com
Abstract This investigation was conducted to establish whether guinea-pig trinitrobenzene sulfonic acid (TNBS)-colitis was associated with a change in the number of neurones of the myenteric plexus, and, if so, whether select subpopulations of neurones were affected. Total neurones were quantified with human (Hu) antiserum, and subpopulations were evaluated with antisera directed against choline acetyltransferase, nitric oxide synthase, calretinin, neuronal nuclear protein or vasoactive intestinal peptide (VIP). Colitis was associated with a loss of 20% of the myenteric neurones, most of which occurred during the first 12 h past-TNBS administration. During this period, myenteric ganglia were infiltrated with neutrophils while lymphocytes appeared at a later time-point. The neuronal loss persisted at a 56-day time-point, when inflammation had resolved. The decrease in myenteric neurones was not associated with a decrease in any given subpopulation of neurones, but the proportion of VIP-immunoreactive neurones increased 6 days following TNBS administration and returned to the control range at the 56 days. These findings indicate that there is an indiscriminant loss of myenteric neurones that occurs during the onset of TNBS-colitis, and the loss of neurones may be associated with the appearance of neutrophils in the region.
Inflammatory bowel disease (IBD) and animal models of intestinal inflammation are associated with changes in gut function and sensitivity.1,2 These changes likely involve inflammation-induced alterations in the enteric neurones that regulate intestinal motility and secretion and the extrinsic sensory neurones that innervate the gut. Both humoral and cell-mediated inflammatory responses are associated with increased excitability of submucosal and myenteric neurones,3–5 and synaptic facilitation.4,5 Other alterations that have been reported in various models of intestinal inflammation include changes in neurotransmitter content of the intestine and changes in neurone numbers.6–12
Intestinal inflammation induced by trinitrobenzene sulfonic acid (TNBS) involves a cell-mediated immune response consistent with a type I helper T-lymphocyte infiltration.13–17 This inflammation is centred within the lamina propria but immune cell infiltrate extends transmurally. We have recently used TNBS-induced colitis in guinea-pigs to investigate changes in the neural circuitry of the enteric nervous system (ENS).5,18,19 On the basis of their electrical properties, guinea-pig enteric neurones are divided into two subclasses, AH and S neurones.20,21 AH neurones are phasic and have a prolonged after hyperpolarization, and serve afferent and interneuronal functions. S neurones are often tonic, receive fast excitatory synaptic input and function as interneurones and motor neurones. TNBS-colitis in the guinea-pig is accompanied by changes in the electrical properties of AH neurones that include increased spontaneous action potentials, a decrease in the after hyperpolarization and increased excitability. In these animals, fast synaptic potentials are significantly more common in AH neurones and are larger in S neurones of the myenteric plexus.5
The guinea-pig is the best understood model of the electrical, morphological and neurochemical properties of enteric neurones; therefore, the current investigation was undertaken to evaluate the structural and neurochemical characteristics of myenteric neurones in the TNBS-inflamed guinea-pig distal colon, where functional changes in enteric neurones have been thoroughly characterized. Human (Hu) immunoreactivity was used to compare the number of neurones in control and inflamed preparations, and double label immunohistochemistry was used to evaluate the proportions of neurones that expressed various markers of select subpopulations of myenteric neurones. The findings of these studies indicated that there was a non-discriminate loss of neurones in the myenteric plexus of TNBS-inflamed tissue. Therefore, the time course of neuronal loss was investigated and correlated with an evaluation of the types of immune cells that were in the vicinity during the window of time when neuronal cell loss occurred. The findings suggest that most neuronal cell loss occurs during the first 12 h, which corresponds to a time span when neutrophils, but not lymphocytes, are present in the region.
Materials and methods
All methods used in this study were approved by the University of Vermont Animal Care and Use Committee. Adult guinea-pigs (Charles River, Montreal, Canada) of either sex, weighing 250–350 g, were housed in metal cages with soft bedding. The animals had access to food and water ad libitum and maintained at 23–24 °C on a 12 : 12 h light–dark cycle.
In order to generate inflammation in the distal colon, guinea-pigs were anaesthetized with isoflurane (induced at 4%, maintained on 1.5% in oxygen) and 0.3 mL of TNBS (25 mg mL−1) in 30% ethanol was delivered into the lumen of the colon through a polyethylene catheter inserted rectally 7 cm proximal to the anus. Control animals remained naive until tissue collection. Animals were maintained in a controlled environment for 2, 6 or 12 h or 1, 6 or 56 days after TNBS administration. At the time of tissue collection, animals were deeply killed with an overdose of isoflurane and exsanguinated. The severity of colitis was assessed by macroscopic colonic damage scoring as described previously.19
The distal colon, identified as the part of the colon between the hypogastric flexure and the pelvic brim, was removed and placed in iced HEPES buffer (in mmol L−1: NaCl, 121; KCl, 5.9; CaCl2, 2.5; MgCl2, 1.2; NaHCO3, 25; NaH2PO4, 1.2 and glucose, 8; pH 7.4; all from Sigma, St Louis, MO, USA). The distal colon was opened along the mesenteric border and pinned flat. These preparations were fixed for 2–16 h at 4 °C in 0.1 mol L−1 phosphate-buffered saline (PBS; 0.1 mol L−1 pH 7.4) containing 4% paraformaldehyde, and 0.2% picric acid. Following three washes with PBS, the mucosa, submucosa and circular muscle of the colon were removed with forceps exposing the myenteric plexus on the longitudinal smooth muscle.
Immunohistochemistry of whole-mount preparations
Some dissected preparations were washed with PBS and incubated for 2 h at room temperature with PBS containing 0.5% Triton X-100 and 4% normal goat serum. This solution was removed, and the sections were incubated overnight at room temperature in PBS containing 4% normal goat serum, 0.5% Triton X-100, a 1 : 200 dilution of mouse antihuman neuronal protein HuC/HuD antiserum (Molecular Probes, Eugene, OR, USA) and one of the following antisera: 1 : 500 dilution of rabbit antinitric oxide synthase (NOS; Santa Cruz Biotechnology, Santa Cruz, CA, USA), 1 : 500 dilution of sheep anti-NOS (Chemicon International, Inc., Temecula, CA, USA), 1 : 500 dilution of rabbit anticalretinin (Chemicon International, Inc.), or 1 : 100 dilution of rabbit antivasoactive intestinal peptide (VIP; Dr John Walsh, UCLA, CURE, code no. 7913). Following three 15 min washes with PBS, these tissue sections were incubated with a 1 : 400 dilution of indocarbocyanine (Cy3)-conjugated goat antirabbit antiserum (Jackson ImmunoResearch, West Grove, PA, USA) and a 1 : 100 dilution of fluorescein isothiocyanate (FITC)-conjugated goat antimouse antiserum (Jackson ImmunoResearch) in PBS containing 0.5% Triton X-100 for 2 h.
Some tissue samples were washed with PBS and incubated for 2 h at room temperature with PBS containing 0.5% Triton X-100 and 4% normal horse serum. This solution was removed, and the sections were incubated overnight at room temperature in PBS containing 4% normal horse serum, 0.5% Triton X-100, a 1 : 200 dilution of mouse anti-Hu antiserum and a 1 : 100 dilution of goat anticholine acetyltransferase (ChAT; Chemicon International, Inc.), washed and incubated for 2 h in PBS containing 0.5% Triton X-100, a 1 : 500 dilution of Cy3-conjugated donkey antigoat antiserum (Jackson ImmunoResearch) and a 1 : 200 dilution of FITC-conjugated donkey antimouse antiserum (Jackson ImmunoResearch).
Another group of tissue samples were washed with PBS and incubated for 2 h at room temperature with PBS containing 0.5% Triton X-100 and 4% normal goat serum. This solution was removed, and the sections were incubated overnight at room temperature in PBS containing 4% normal goat serum, 0.5% Triton X-100, a 1 : 400 dilution of mouse antineuronal nuclei (NeuN) antiserum (Chemicon International, Inc.), washed, and incubated for 2 h in PBS containing 0.5% Triton X-100, a 1 : 400 dilution of Cy3-conjugated goat antimouse antiserum. After washing, the tissue samples were incubated overnight at room temperature in PBS containing 4% normal goat serum, 0.5% Triton X-100, a 1 : 200 dilution of mouse anti-Hu antiserum, washed, and incubated for 2 h in PBS containing 0.5% Triton X-100, a 1 : 100 dilution of FITC-conjugated goat antimouse antiserum.
Following the last incubation period, all tissue samples underwent three 15 min washes with PBS, a quick dip in water and were mounted on 3-aminopropyltriethoxysilane (APTEX)-coated (Sigma Chemical Co.) slides and coverslipped with Citifluor (Citifluor Limited, London, UK).
Mounted slides were examined on an Olympus AX70 fluorescence photomicroscope (Olympus America, Inc, Melville, NY, USA). Filter sets included the following: for Cy3, excitation 510–550 nm, emission, 590 nm; for FITC, 470 nm excitation, 525 nm emission. Images were captured with an Optronics MagnaFire digital camera (Optronics, Goleta, CA, USA) attached to the microscope. Images were cropped in Adobe Photoshop (Adobe Systems, San Jose, CA, USA) with minimal alteration (minor adjustments to brightness and contrast), and transferred to Microsoft Powerpoint for construction of figure sets.
Analysis of myenteric immune infiltrate
Whole-mount preparations of the longitudinal muscle/myenteric plexus were generated as described above. Several segments of distal colon, 0.5 cm in length, were stacked and sandwiched between pieces of abdominal skeletal muscle. This block was fixed in 10% neutral-buffered formalin (pH 7.4), paraffin-embedded and sections were cut on a microtome at a thickness of 5 μm. Sections were stained with haematoxylin and eosin and coverslipped. Sections that contained myenteric plexus were analysed for immune cell infiltrate. Neural tissue was centred in a 0.23 mm2 field of view (400×) and the number of neutrophils, eosinophils and lymphocytes, assessed on morphological and histochemical bases, were counted. These determinations were made at three random locations in each section. Three tissue sections, at least 50 μm apart to ensure no replicate counts, were assessed from each animal.
Time course of TNBS-induced inflammation
A single administration of TNBS/EtOH in the guinea-pig distal colon caused regional inflammation that was characterized by ulceration, hyperaemia, adhesions and oedema, and was similar to that described in previous reports.5,19 Macroscopic damage scores revealed that the colon of TNBS-treated animals was significantly damaged as early as 2 h following TNBS administration. Furthermore, damage to the colon was greatest at 12 h following TNBS administration, remained significantly damaged at 6 days post-TNBS but returned to levels that were not significantly different than controls (56 days after TNBS; Fig. 1).
Loss of myenteric neurones in TNBS-treated animals
The numbers of Hu-immunoreactive (IR) neurones were quantified in whole-mount longitudinal muscle myenteric plexus preparations of distal colon from control animals and from animals 6 h, 12 h, 24 h, 6 days and 56 days after intracolonic administration of TNBS in ethanol (Figs 2 and 3). The mean number of neurones per ganglion was significantly reduced by approximately 15% in tissue from animals (12 and 24 h) after TNBS compared with controls. The number of neurones per ganglion was unchanged, relative to healthy control values, in tissue from animals 24 h after treatment with EtOH alone (102.4 ± 2.3 neurones per ganglion). At 6 days post-TNBS the number of neurones per ganglion was 20% less than controls and remained at this level in tissue from animals 56 days post-TNBS, after recovery from inflammation.
Further analysis of the neuronal loss was conducted at the 6-day time-point because neuronal loss had peaked at this time. The frequency distribution of number of neurones per ganglion was symmetrically shifted to the left in inflamed tissue (Fig. 3A; P < 0.05; chi-squared test), suggesting that the decline in neurones per ganglion was due to neuronal loss rather than a restructuring resulting in smaller ganglia. In further support of this, the mean area of individual ganglia and frequency distribution of ganglion area were not altered in inflamed tissue (Fig. 3B,C; P > 0.05; Student's t and chi-squared tests). Possible changes in the elasticity of the tissue did not affect these results because neither the mean number of ganglia in a low magnification field of view (4× objective), nor the number of ganglia situated in a transverse plane were altered in the inflamed tissue relative to control (Fig. 4).
Loss of myenteric neurones is not restricted to a neurochemically defined subpopulation
Subpopulations of myenteric neurones in the guinea-pig distal colon can be characterized by immunoreactivity for neurochemical markers.22 Antisera directed against five neurochemical identifiers were used to evaluate myenteric neuronal subpopulations in control and inflamed tissues at 6 days post-TNBS administration: anti-ChAT was used to identify cholinergic neurones; anti-NOS was used to identify nitrergic neurones, including inhibitory motor neurones; anti-NeuN marker was used to identify intrinsic primary afferent neurones; anticalretinin was used to identify longitudinal muscle excitatory motor neurones and ascending interneurones; and anti-VIP was used to identify a subset of inhibitory motor neurones. When evaluated on the bases of the number or proportion of neurones per ganglion, no changes were detected in the neuronal subpopulations IR for ChAT, NOS, NeuN or calretinin (Table 1; Fig. 5). On the contrary, a significant increase in the number and proportion of neurones that were VIP-IR was detected (Table 1; Fig. 5) and most (76.7 ± 1.2%) of these neurones were also NOS-IR. At the time-point of 56 days post-TNBS administration, the VIP subpopulation had returned to the control level (0.10 ± 0.03 VIP-IR neurones per ganglion; 0.12 ± 0.04% VIP-IR neurones). No obvious differences in the density of IR nerve fibres were observed for any of these markers were observed in TNBS-treated vs control preparations.
Table 1. Number and proportion of neurochemically defined neurones per myenteric ganglion
Number of neurones per ganglion
Proportion of neurones per ganglion
The mean number of neurochemical immunoreactive cell bodies per ganglion and the proportion of neurones per ganglion that display neurochemical immunoreactivity.
*P < 0.05 compared with control, Student's t-test.
Inflammatory infiltrate within the myenteric plexus involves neutrophils
Neuronal loss in the myenteric plexus of inflamed tissue could be caused by a number of mechanisms. Our working hypothesis is that the immune response to intraluminal TNBS involves infiltration of the myenteric plexus, which could lead to immune-neural communication and ultimately cell loss. To investigate what type(s) of cellular infiltrate correspond to the time course of neuronal cell loss, immune cells were enumerated in en face sections of stacked longitudinal muscle–myenteric plexus preparations that were stained with haematoxylin and eosin. Neutrophils, eosinophils and lymphocytes were identified on the basis of morphology and were counted in high magnification fields of view (40× objective) that were centred on myenteric neural tissue (Table 2). Neutrophils were not detected in sections from control animals; however they were detected in inflamed tissue as early as 2 h post-TNBS administration, and their numbers were significantly greater than control at 6- and 12-h time-points (Fig. 6). At 6 days post-TNBS, the number of neutrophils had decreased to <1 per field. Neutrophils were observed within and surrounding myenteric ganglia. Eosinophils were significantly increased in tissue from animals 12 h post-TNBS, but were still limited to about 1 per field. The number of lymphocytes was significantly greater in tissue 24 h post-TNBS. Lymphocytes were seen in the vicinity of neural tissue, but unlike neutrophils, they were rarely observed within ganglia. The differential time courses of immune cell infiltrate suggest that neutrophils are present at the time-points when myenteric neurones are being lost.
Table 2. Time course and type of immune infiltrate in the trinitrobenzene sulfonic acid (TNBS)-inflamed colon
Time after TNBS treatment
The mean (±SEM) number of immune cells in a high magnification field of view centred on myenteric plexus neural tissue.
*P < 0.05 compared with control, anova, Newman–Keuls multiple comparisons test (n = 3–4).
0.07 ± 0.07
0.07 ± 0.07
2.7 ± 2.0
0.10 ± 0.06
0.1 ± 0.1
24 ± 7*
1.2 ± 0.5
1.0 ± 0.4
20 ± 3*
1.3 ± 0.1*
1.9 ± 0.6
22 ± 14
1.0 ± 0.8
7 ± 4*
0.4 ± 0.2
0.07 ± 0.07
0.20 ± 0.08
The purpose of this investigation was to identify what changes occur in the numbers and neurochemical properties of neurones in the myenteric plexus of the TNBS-inflamed guinea-pig distal colon. Our findings indicate that inflammation leads to an indiscriminant loss of about 20% of the myenteric neurones. Most of the neuronal loss occurred during the first 12 h following TNBS administration, and it persisted following recovery from inflammation. The only change that was detected in neurochemical markers tested was an increase in the number and proportion of VIP-IR neurones, which returned to control levels at 56 days after TNBS administration, when inflammation had resolved.
This is not the first study to demonstrate a change in the number of neurones in the inflamed bowel. In Crohn's disease and ulcerative colitis, both neuronal hyperplasia and neuronal degeneration have been reported, but it is difficult to interpret these results because they involved evaluation of cross-sections and the time course of the disease process was unclear (see Ref.9). In experimental colitis decreases in neuronal numbers have been consistently reported. For example, a 50% decrease in the number of neurones per ganglion was detected in cross-sections of dinitrobenzene sulfonic acid (DNBS)-inflamed rat colons at a 24-h time-point.10 In rat TNBS-colitis, neuronal markers were absent in severely inflamed regions, although no effort was made to quantitate these findings.11 In mouse DNBS-induced colitis, a 50% decrease in the number of myenteric neurones was detected in cross-sections while a smaller reduction in cell numbers (40%) was detected in whole-mount tissues.12 On the contrary, in guinea-pig TNBS-ileitis, there was no significant change in the number of neurones per ganglion at time-points ranging from 1 to 14 days post-TNBS administration.23 This may reflect the high degree of TNBS-induced inflammation that occurs in the colon as opposed to the ileum.19,24
In the current study, we evaluated neuronal cell numbers in whole-mount longitudinal muscle–myenteric plexus preparations. From this en face perspective, all of the neurones in a given area can be visualized because they exist as a monolayer in the plane of the myenteric plexus. In guinea-pigs with TNBS-colitis, the number of neurones per ganglion decreased by 20%. This likely represents a 20% loss of myenteric neurones, rather than a change in the structural organization of the ganglia or the plexus, because neither the mean ganglionic area nor the density of ganglia was altered. The decrease in density of neurones per ganglion could result in the appearance of a more dramatic reduction of neurones when viewed in cross-section, and this may explain why more extensive cell loss has been described in other studies.
The neurones of the myenteric plexus include intrinsic primary afferent neurones, intestinofugal afferent neurones, ascending and descending interneurones, and excitatory and inhibitory motor neurones.22,25 These subpopulations can be distinguished by unique combinations of neurochemical markers that they express in the guinea-pig. For example, intrinsic primary afferent neurones express the biosynthetic enzyme for acetylcholine, ChAT and the protein NeuN.26 Several neurochemical markers that label multiple subpopulations of myenteric neurones were examined in this investigation, in combination with Hu to label the entire neuronal population. If inflammation-induced differences were detected in any given broad subpopulation subsequent multiple label studies would have been conducted to resolve which group(s) of neurones were affected. These results suggest that the 20% loss of myenteric neurones was not restricted to a given subpopulation of neurones. This interpretation is consistent with electrophysiological data demonstrating that the proportions of cells belonging to the two major classes of electrically defined enteric neurones, AH and S neurones, are not altered in the inflamed colon.5 Electrophysiological studies also revealed synaptic plasticity in the myenteric ganglia that include an increase in the proportion of AH neurones that receive fast excitatory synaptic input and an increase in the amplitude of fast excitatory postsynaptic potentials in S neurones.5 These forms of synaptic facilitation may involve neuroanatomical reorganization of the myenteric circuitry, and axonal sprouting in response to the loss of neurones.
The only change that was detected in the neurochemical properties of myenteric neurones in the inflamed colon was an increase in the number and proportion of myenteric neurones that were VIP-IR, and this change was transient as the values returned to the control range following recovery from inflammation. It is not yet clear whether this change reflects an increase in the number of neurones synthesizing VIP, an increase in the amount of peptide synthesized by VIPergic neurones, and/or a slowing of the movement of the peptide into the axons, resulting in enhanced immuoreactivity within the neuronal cell bodies. In a study that involved evaluation of VIP-IR nerve fibres in rats with TNBS-colitis, a transient decrease in VIP-IR nerve fibres, particularly in the mucosa, was reported.6 In rectal biopsies of individuals with Crohn's disease, there is an increase in the number of VIP-IR neurones in the submucosal ganglia of non-inflamed regions,7 and an increase in VIP immunoreactivity has also been reported in rectal biopsies of individuals with moderate to severe ulcerative colitis.8 It is possible that VIP plays a protective role in the inflamed gut, as administration of VIP to TNBS-treated mice reduces the severity of colitis.27 Further studies will be necessary to characterize the mechanisms and physiological significance of the increase in VIP-IR neurones that is reported here.
In animal models of intestinal inflammation, the loss of myenteric neurones occurs quite rapidly. In rat DNBS-colitis, myenteric neuronal loss was not yet evident at a 6-h time-point but was evident at 24 h.10 In mouse DNBS-colitis, loss of neurones was evident at the 6-h time-point and markers of neuronal apoptosis were evident as early as 30 min after insult.12 In the current study, while no neuronal loss was detected at 6 h following administration of TNBS, loss of 15% of the neurones was evident at the 12- and 24-h time-points, and 20% neuronal loss was detected at the 6- and 56-day time-points. Therefore, 75% of the neuronal loss caused by TNBS-induced inflammation occurred within the first day.
The mechanism of neuronal loss in these animal models is not completely understood, but it appears to involve infiltration of neutrophils, and possibly eosinophils. Administration of steroids, to attenuate the inflammatory response, significantly reduces neuronal cell loss in rat DNBS-colitis.10 However, this model differs from the guinea-pig TNBS-colitis model in that eosinophils were the primary infiltrating immune cells during the period of neuronal loss. In mouse DNBS-colitis, administration of an antineutrophil serum reduces inflammation and attenuates the loss of myenteric neurones.12 In the present study, neutrophils were present in the vicinity of the myenteric plexus as early as 2 h post-TNBS. The neutrophil infiltration peaked at 6 h and remained at these levels through the 24-h time-point, whereas lymphocytes appeared at the 24-h time-point when most of the neuronal loss had already occurred. Importantly, invading neutrophils penetrated the ganglia whereas lymphocytes were only detected in the surrounding region. This is consistent with guinea-pig TNBS-ileitis, where T cells IR for CD4 and CD8 were observed surrounding, but rarely within, myenteric ganglia.23 Interestingly, while these data support the involvement of neutrophils in neuronal loss, neutrophils do not appear to contribute to the mucosal inflammatory response in TNBS-colitis.28
In conclusion, TNBS-colitis in the guinea-pig results in a rapid and permanent loss of one in five myenteric neurones. Based on neurochemical coding studies, it appears that the neuronal loss does not occur within a given subpopulation of neurones. As neutrophils are the primary infiltrating immune cell at the time of neuronal loss, it is likely that they are associated with neuronal cell death in this model.
The authors are grateful for the insights and contributions of Dr Edward Parr, Ms Sophie Griswold, Mr Eric Krauter, Mr Andrew Tinsley and Ms Elice Brooks. This work was supported by NIH grants NS26995 and DK62267 (GMM), and P20 RR-016435 from the COBRE program of the National Center for Research Resources, and a grant from the Crohn's and Colitis Foundation of Canada (KAS, GMM). KAS is a Medical Scientist of the Alberta Heritage Foundation for Medical Research.